Calibration Data Transfer

The V3 way of calibrating a reactor involved running it in Excel and then exporting factors from Excel to the relevant Petro-SIM case. V4 replaces that with new mechanisms. What are they and when and why do you use them? Let's explore the why first.

V4 calibrates inside Petro-SIM so can you have one case that you use for calibration and prediction?

You can but the reality is that sites often end up with many Petro-SIM cases covering different scopes and uses. If you are just getting started with Petro-SIM or are a SIM customer interested primarily in the reactor, one case for all uses will probably work. As your use grows and particularly as you start building cases that represent multiple units, including calibration data in all cases will quickly become unwieldly, making it hard to keep things consistent.

What transfer methods does Petro-SIM have?

There are two basic methods:

  • XML files
  • Petro-SIM database

The information generated is the same in both cases and includes all the measurement and results data from a calibration. It is not complete case information. With XML, each export generates a new time-stamped dataset XML written to the User Support Files location on your machine, where this is set in Preferences. With the database, each dataset is written to your connected database.

XML Transfers

Suppose you have two process units (an FCC and a Reformer) that get calibrated by the relevant unit engineer. You have another flowsheet that includes both these reactors that you use for studies. With XML the data flows are:

Here we assume that the unit calibrations have been updated from time to time, with the unit engineer exporting each as a new dataset xml file. These get imported to the Refinery case and as part of this import you can filter how much information is read in.

If you want to share calibrations between users, there are two primary mechanisms you can use:

  • XML files are saved to network folders or workspaces, with the calibration user informing other team members when new data is available
  • XML files are emailed to other team members

Database Transfers

The data flows are the same here but the storage mechanism for the data is quite different: information resides in the Petro-SIM relational database allowing Petro-SIM to provide some additional functionality. The data flows look like:

Here we have chosen to make each new calibration a revision of the previous one, making it easier for us to see a calibration history. We see other advantages as we import the data to the refinery case in Petro-SIM:

  • The import process scans the database for all available data, showing us only datasets that include the relevant object type. This reduces the need to rely on file naming conventions!
  • Petro-SIM gives you an option to remember which dataset supplied the calibration. It then checks each time you load the case to see if a new calibration is available, alerting you if anything is found.

We have a distributed team: will Databases work for us?

The Petro-SIM database can be your case repository as well as your calibration data repository. It is important to make a distinction between these uses: storing large cases with today's technology and typical corporate WAN speeds is not really practical and only makes sense if the project team are all based in the same physical location. Calibration data is significantly smaller however and is feasible on most networks or over VPN connections. Even when you scan the database for available datasets, that search query is executed on the database server and only the query results have to be transferred.

We recommend you experiment on your network before making any final decision on transfer techniques: WAN and database speeds do vary and you may find you can store cases as well as the calibration data.

Tracking where calibrations come from

Petro-SIM maintains some information in each case so you can see when an object was calibrated and where the data came from. This happens each time you calibrate an object or import a calibration dataset from XML or from the database. You can view what gets saved on the Source page on the Calibration Factors tab of each reactor. Petro-SIM tracks:

  • Calibration Date - the time when "Accept" was clicked
  • Lab Date - where your calibration pulls data from a process historian, this will be the process end time used.
  • Calibrated By - name of the user who ran the calibration
  • User ID - their Windows login id
  • Computer Name - the machine they calibrated on
  • Case Name - the Petro-SIM case name in which calibration occured (may also have a DB connection row if the case is a database case)
  • Data Set - this is the XML or DB file from which you imported the calibration

Case names might often be on the individual users machine and not on a shared resource: the information here should give you enough to locate the original files!